marco at simplex.nl
Sat Jan 13 11:27:49 PST 2001
From: "Lourens Veen" <jsr at dds.nl>
> Looks nice, although I think that the current very-beta Tarkin sources
> do more or less the same. They use a 3D wavelet transform, which is
> probably better than differential encoding in the time domain. You can
> grab the current rather crude IMHO Tarkin code here:
If anyone is interested in some references on 3D coding give a shout. I searched
around a bit for papers on 3D and 2+1D encoders, but my papers directory is in
complete dissarray. Ill just give a pointer to a paper on the most recent (and
most successfull) implementation of 3D wavelet coding Im aware of, hope someone
finds it usefull...
http://www.microsoft.com/CHINA/research/downloads/g-imedia/spli/IMG-3.pdf its a
2+1D approach with the twist that they do the 1D transform along motion paths,
with pretty impressive results.
I have papers a lot more applicable to the present state of Tarkin, as I
understand it, lost in a big pile... can anyone suggest a good PDF indexing
program? Acrabat Catalog is crap, especially given that a good amount papers can
not be text searched.
--- >8 ----
List archives: http://www.xiph.org/archives/
Ogg project homepage: http://www.xiph.org/ogg/
To unsubscribe from this list, send a message to 'vorbis-dev-request at xiph.org'
containing only the word 'unsubscribe' in the body. No subject is needed.
Unsubscribe messages sent to the list will be ignored/filtered.
More information about the Vorbis-dev